回转支承专家系统的推理策略及模糊推理技术

通过建立回转支承专家系统，研究机械设计专家系统的结构，问题求解策略及模糊推理等技术。提出将知识库中的知识分层从而实现推理中的动态任务分解。推理机通过结合知识库中最高层次的元规则产生日程表，确定当前任务，然后动态调用下属层次的元规则，查找并使领域知识以完成分解了的任务。推理时利用“黑板”使同层次推理机构间知识共享，从而可完成复杂问题的求解。处理不确定的性问题时，证据可信度的得到也是依靠任务分解，由下

Inference Strategy and Fuzzy Inference for the Expert System of Slewing Ring Bearings

In the course of setting up the expert system of slewing ring bearings (SRBES),the architecture of machine design expert system,problem solving - strategy and the fuzzy reasoning have been studied in the paper.Knowledge in its base was layered,so as to realize the dynamic task decomposition in reasoning. By means of the top meta rules in knowledge base the inference engine produces the routine schedule, determines the current solving subgoal,dynamically call the lower layered meta rules and does work out the decomposed task with domain knowledge. In the course of reasoning, 'blackboard' has been employed to afford those inference machinery at the same layer sharing all the information in order to solve the complex problems. For uncertain problem, the values of certainty factor of the evidence are also obtained by submachinery,relying on the task decomposition. The revision of the certainty factor of conclusion is worked out by calculation module of certainty factor under the control of meta rules.